Re: Bug with fcntl(n, F_SETOWN, getpid) with NPTL 0.34

Roland McGrath wrote:
> > It is incorrect for a program to use the value of getpid() as the
> > argument to F_SETOWN.  It should be using the value returned from the
> > gettid() system call (tid (current->pid) != pid (current-tgid)).  (Yes
> > the naming is very confusing).  Unfortunately this is not made easy
> > (gettid() is not a libc function), and the man page is no longer
> > correct when it says process id.
> 
> This is incorrect.  The kernel is what needs to be fixed.

I agree the kernel could be changed.  But which change do you have in mind?

Changing fasync to deliver signals to a process instead of a thread
breaks programs which use sigwaitinfo() as their per-thread event
queue mechanism (e.g. to track dnotify events, or to guarantee event
ordering from async futexes).

There was a recent tiny kernel thread on this topic.
As usual I may be the only person in the world who actually uses these
capabilities ;)

-- Jamie